Mistakes rarely manifest as split second incorrect decisions. Sure, in Pharmacy, dishing out the wrong medication often involves a short, momentary, lapse of concentration where the deed occurs. But many other errors, as such, occur through misreading a situation over a period of time.

We often hear of a patient whose diagnosis was 'missed'. Often these diagnoses are not glaringly obvious. The picture changes over time; it's when there is a level of clinical dissonance the doc needs to rethink their diagnosis. There are many possible outcomes for any given constellation of presenting complaints.

These thoughts wafted into my conscious early this morning, about halfway through a cycling race. In previous cycling races I've been in, when the breakaway jumps off the front of the race, either you see it coming or can react fast enough to go with it. From this piece of information and the fact that my mind was wandering to medical errors, pharmacy and planning a blog post, all whilst racing, you would assume, right then and there, I missed the break.

But today's race was in the hills. Plenty of time to think whilst two young fellows managed to make the weaker riders do all the work, and one by one, we all ran out of puff. Then could they ride away to the victory, we other riders turning the pedals and going nowhere as they launch off to victory.

So, about the time I'm sitting on the front, mind wandering into fairyland and using my energy to maintain a decent pace, something clicks. This is where the mistake happens. For me, it was a cascade of errors; taking a turn at the wrong point, staying on the front too long, burning up energy earlier than necessary, not really concentrating of when to save or burn energy.

And, in fact, the vast majority of medical and pharmacy errors appear to occur the same way. As I discovered during my PharmSchool research project, the literature describes it as a Swiss Cheese model. Only when slices of Swiss Cheese line up, can a certain trajectory of error pass through.

In pharmacy, this is characterised by multiple procedures, checks and double checks for each 'script. In medicine immediate analysis occurs by peer-reviewed diagnostic and management plans; and retrospectively by Morbidity and Mortality(M&M) conferences.

In these meetings, clinicians analyze situations that are considered to be less than ideal, either for the outcome of the process. My previous impression was that M&M was a bit of a witch-hunt; a finger-pointing exercise. They're not.

In the ideal M&M meeting, the scenario is reproduced objectively, and the other clinicians challenged to think;

"What would I do here? What am I concerned about? What information do I need to gather?"

At a particular time point.

Decision making becomes much easier if you already have all of the information, and a 20:20 retrospectoscope. In real life, decisions are made much earlier, borne of necessity with the information available at the time. Often cases in the M&M meetings have a "swing point", where suddenly everything comes together, just like an episode of House MD. In M&M cases, that point is "too late". In near misses, that point is "just in time". In most cases, the ones that are well handled, that point happens well before the critical time. Long before the break has gone and the teenagers have ridden everyone to the ground.

Mistakes happen. The best we can do is to learn from them, analyze them, and try our level best not to repeat ours or others. But they happen. That's medicine and that's life.